DIGITAL FILTERS FOR ROUTINE DATA REDUCTION
Abstract
Digital tapes produced from analog field data often carry over severe noise problems. A batch data processing capability should therefore be complemented by digital filters which can remove noise routinely. This paper describes a set of filters developed for use with Nuclear Weapons Effects Division standard data processing codes. Emphasis in assembly of the set was placed on accuracy of signal retention and on adaptability to general purpose computers. Part I discusses data which is sampled at equally spaced times, the use of sample numbers to designate time positions, and the requirement for use of a sampling frequency at least eight times greater than the highest expected signal frequency. Parts II, III, and IV document the filters of the three subroutines presently being used for noise removal. The first subroutine eliminates randomly spaced single-sample 'spikes' through the application of inequality conditions to data. The second subroutine provides frequency filters for the removal of zero drift and single noise frequencies. Low-pass frequency filters are available in the third subroutine. Part V discusses the ten example plates used to demonstrate the filter set's effectiveness and speed of operation.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1970
- Accession Number
- AD0705571
Entities
People
- Hendrick D. Carleton